Thorsten Uthmeier & Jürgen Richter

Chapter 14
probably transported while embedded in sediment.
On a slope like this, running water a�er
heavy rainfall would have probably had too
much energy to be responsible for a transportation
that did not severely affect the finds.
2. The assemblage of level VI/13 represents either
a short phase of down slope movement, whereas
level VI/14 was transported earlier and levels
VI/11 and VI/12 later, or consists of the main
part of a concentration that was situated some
metres above today’s topographical position
and transported as an entity.
As nearly one half of all flakes were completely (4
cases) or partly (19 cases) covered by cortex on their
dorsal surfaces (Fig. 14-21), it is possible to reconstruct
aspects of raw material procurement strategies.
A total of 14 raw material units include artefacts
that display the remains of a chalky cortex originally
covering the nodules (Fig. 14-22). This points to primary
sources, possibly situated in the Bodrak Valley
some 6 km from the site. Local outcrops in the
immediate vicinity of Kabazi Mountain which are
known today were not accessible at the time level
VI/13 accumulated: according to the elevation of the
3 rd terrace of the Alma, which is situated some metres
below the excavated area, they were covered by
sediments until the river cut deeper in the landscape
during OIS 3. Four raw material units had a thin,
hard cortex before the flaking started, and were taken
from secondary sources. As the cortex is not rolled,
and therefore not transported by water, the material
might have been collected near primary sources after
raw material eroded out of the chalky limestone.
Finally, there are two raw material units that show
an opportunistic raw material procurement strategy.
One (Fig. 14-18, 6) is a single piece that resulted from
frost cracking of a thin plaque�e. The other item (Fig.
14-16, 7) is a patinated flake reused by Neanderthals
a�er it had been exposed to sunlight and weathering
for some time. For most raw material units, the
amount of cortex on the blanks was not sufficient to
reconstruct the original shape of raw nodules (Fig.
14-23). For those that allowed hypothesis in this regard,
the flaking started with round (1 case), round
and flat (3 cases), and flat (2 cases) nodules, as well
as with a plaque�e. To summarize, no preferences
for distinct shapes of raw pieces could be observed,
possibly due to the low total number of artefacts in
each raw material unit.
Transformation Analysis at Kabazi II, Levels VI/11-14
Results of transformation analysis:
a critical discussion
The results of transformation analysis are presented
in three ways: in a diagram showing the frequency
of transformation sections (Fig. 14-24), in a table
documenting the frequency of blanks in each raw
material unit which were used as defining criteria
for the classification of transformation sections (Table
14-1), and in a diagram that depicts the chaîne
opératoire reconstructed for every raw material unit
as a flow chart (Fig. 14-25 and 14-26). It has already
been mentioned that 11 raw material units were single
pieces. Of these, five are blanks (Table 14-1 and
Fig. 14-25: RMUs 1, 12, 8, 22, 23 classified as “Bw”),
and one is a core (RMU 4, classified as “Cw”). Four
tools also proved to be single pieces (“Tw”). Two
are simple side scrapers (Table 14-1 and Fig. 14-25:
RMUs 3, 16), one is a double side scraper (Table 14-1
and Fig. 14-25: RMU 18), and one is a surface shaped
bifacial scraper (Table 14-1 and Fig. 14-25: RMU 19).
Another single piece (Table 14-1 and Fig. 14-25: RMU
11) is a ventral lateral sharpening flake that removed
the basal part of the le� working edge of a quite
large unifacial tool. According to the definitions for
classes of transformation sections (Uthmeier 2004a),
it is the isolated end of a tool (“Ei”), indicating that a
modified piece was rejuvenated on the site and then
taken out of the excavated area. Once again, it has to
be stressed that incomplete raw material units might
be the result of natural artefact transport, unsuccessful
a�empts to separate different archaeological levels,
or transport by humans. Since the site formation
process of level VI/13 was characterised by post-depositional
movements of artefacts or sediments, none
of these hypotheses can be excluded. As there is no
clear pa�ern visible, but a mixture of cores, blanks,
simple tools and surface shaped tools, the authors
tend to believe that human transport of ready made
tools or prepared cores played only a minor role in
the presence of single pieces.
In total, 11 raw material units represent flaking
activities conducted within the excavated area,
and therefore on the site (Fig. 14-25; 14-26). However,
there are still uncertainties to what extent the
amount of flaking reconstructed by transformation
analysis really correspond to prehistoric reality. This
is especially relevant for two short transformation
sections (Fig. 14-27) covering only one phase of the
formal chaîne opératoire. They are suspicious because
additional artefacts, probably hidden in other levels,
might essentially change the classification of transformation
sections. One unit (Table 14-1 and Fig. 14-26:
RMU 24) consists of two chips that fulfil the definitions
for the correction of a core (or preforms: “Cc”).
239